History | View | Annotate | Download (21.8 kB)
error: Strip trailing '\n' from error string arguments (again)
Commit 6daf194d and be62a2eb got rid of a bunch, but they keep comingback. Tracked down with this Coccinelle semantic patch:
r expression err, eno, cls, fmt; position p; @@ (...
r
qdev: drop extra references at creation time
qdev_free and qbus_free have to do unparent+unref, because nobody elsedrops the initial reference (the one included by object_initialize)before them.
For device_init_func and do_device_add, this is trivially correct,...
qdev: add reference count to a device for the BusChild
Each device has a reference through the BusChild. This referencewas not accounted for, add it now.
Reviewed-by: Andreas Färber <afaerber@suse.de>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>...
qdev: move deletion of children from finalize to unparent
A device will never be finalized as long as it has a reference fromother devices that sit on its buses. To ensure that the referencesgo away, deassociate a bus from its children in the unparent callback...
qdev: move unrealization of devices from finalize to unparent
Similarly, a bus holds a reference back to the device, and this willprevent the device from going away as soon as this reference is countedproperly. To avoid this, move the unrealization of devices to the...
qdev: add reference for the bus while it is referred to by the DeviceState
Now that the unparent callbacks are complete, we can correctly accountmore missing references.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
qdev: inline object_delete into qbus_free/qdev_free
We want object_delete to disappear, and we will do this one class at atime. Inline it for the qdev case, which we will tackle first.
qdev: remove duplication between qbus_create and qbus_create_inplace
Move the common part to qbus_realize.
Acked-by: Andreas Färber <afaerber@suse.de>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
qdev: change first argument of qbus_create_inplace to void *
Make it clear that no BUS macro is needed in the callers (in fact itwouldn't work because the object has not been initialized yet with theright class).
Suggested-by: Andreas Faerber <afaerber@suse.de>...
qdev: Prepare "realized" property
Introduce the QOM realizefn suggested by Anthony.Detailed documentation is supplied in the qdev header.
For now this implements a default DeviceClass::realize callback thatjust wraps DeviceClass::init, which it deprecates....
qdev: Fold state enum into bool realized
Whether the device was initialized or not is QOM-level information andcurrently unused. Drop it from device. This leaves the boolean state ofwhether or not DeviceClass::init was called or not, a.k.a. "realized"....
qdev: add qbus_reset_all
Make all static TypeInfos const
Since 39bffca2030950ef6efe57c2fac8327a45ae1015 (qdev: register alltypes natively through QEMU Object Model), TypeInfo as used inthe common, non-iterative pattern is no longer amended with informationand should therefore be const....
qdev: Don't assume existence of parent bus on unparenting
Commit 667d22d1ae59da46b4c1fbd094ca61145f19b8c3 (qdev: move bus removalto object_unparent) made the assumption that at unparenting timeparent_bus is not NULL. This assumption is unjustified since...
Merge branch 'master' of git://git.qemu.org/qemu into qom-cpu
Adapt header include paths.
Signed-off-by: Andreas Färber <afaerber@suse.de>
qdev: Coding style fixes
Add missing braces and break lines larger than 80 chars.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>Signed-off-by: Andreas Färber <afaerber@suse.de>
qdev-properties.c: Separate core from the code used only by qemu-system-*
This separates the qdev properties code in two parts: - qdev-properties.c, that contains most of the qdev properties code; - qdev-properties-system.c for code specific for qemu-system-*,...
softmmu: move include files to include/sysemu/
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
qapi: move include files to include/qobject/
net: reorganize headers
Move public headers to include/net, and leave private headers in net/.Put the virtio headers in include/net/tap.h, removing the multiple copiesthat existed. Leave include/net/tap.h as the interface for NICs, andnet/tap_int.h as the interface for OS-specific parts of the tap backend....
qdev: qdev_create(): use error_report() instead of hw_error()
hw_error() is specific for fatal hardware emulation errors, not forinternal errors related to the qdev object/class abstraction or objectinitialization.
Replace it with an error_report() call, followed by abort()....
qdev: move bus removal to object_unparent
Add an ObjectClass method that is done at object_unparent time. Itshould remove any backlinks to the object in the composition tree,so that object_delete will be able to drop the last reference andfree the object....
qdev: simplify (de)allocation of buses
All conditional deallocation can now be done with object_delete.Remove the @qom_allocated and @glib_allocated fields; replace the latterwith a direct assignment of the @free function pointer.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>...
qdev: Split up header so it can be used in cpu.h
Header file dependency is a frickin' nightmare right now. cpu.h tendsto get included in our 'include everything' header files but qdev alsoneeds to include those headers mainly for qdev-properties since it knows...
qdev: allow multiple qdev_init_gpio_in() calls
Allow multiple qdev_init_gpio_in() calls for the one device. The first call willdefine GPIOs 0-N-1, the next GPIOs N- ... . Allows different GPIOs to be handledwith different handlers. Needed when two levels of the QOM class heirachy both...
Merge branch 'trivial-patches' of git://github.com/stefanha/qemu
vl.c: convert *vga_enabled functions to QOM
And get rid of qdev_exists().
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
qdev: kill bogus comment
When the DeviceInfo code was removed, the comment describingqdev_subclass_init() was left in the code by mistake. Remove it.
Cc: qemu-trivial@nongnu.orgSigned-off-by: Eduardo Habkost <ehabkost@redhat.com>Signed-off-by: Stefan Hajnoczi <stefanha@gmail.com>
scsi, pci, qdev, isa-bus, sysbus: don't let *_get_fw_dev_path return NULL
Use g_strdup rather than strdup, because the sole caller(qdev_get_fw_dev_path_helper) assumes it gets non-NULL, and dereferencesit. Besides, in that caller, the allocated buffer is already freed with...
qom: object_delete should unparent the object first
object_deinit is only called when the reference count goes to zero,and yet tries to do an object_unparent. Now, object_unparenteither does nothing or it will decrease the reference count.Because we know the reference count is zero, the object_unparent...
net: Remove VLANState
VLANState is no longer used and can be removed.
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>Reviewed-by: Laszlo Ersek <lersek@redhat.com>
qdev: fix use-after-free in the error path of qdev_init_nofail
From Markus:
Before:
$ qemu-system-x86_64 -display none -drive if=ide qemu-system-x86_64: Device needs media, but drive is empty qemu-system-x86_64: Initialization of device ide-hd failed...
Merge remote-tracking branch 'afaerber-or/qom-next-2' into staging
qdev: Remove qdev_prop_set_defaults
Instead, qdev_property_add_static can set the default.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>Signed-off-by: Andreas Färber <afaerber@suse.de>
qdev: Use wrapper for qdev_get_path
This makes it easier to remove it from BusInfo.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>[AF: Drop now unnecessary NULL initialization in scsibus_get_dev_path()]...
qdev: Move SysBus initialization to sysbus.c
TYPE_SYSTEM_BUS will be local to hw/sysbus.c, so move existing referencesto main_system_bus and system_bus_info there.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>Signed-off-by: Andreas Färber <afaerber@suse.de>
qdev: Convert busses to QEMU Object Model
This is far less interesting than it sounds. We simply add an Object to eachBusState and then register the types appropriately. Most of the interestingrefactoring will follow in the next patches.
Since we're changing fundamental type names (BusInfo -> BusClass), it all needs...
qdev: Connect busses with their parent devices
This makes SysBus part of the root hierarchy and all busses children oftheir respective parent DeviceState.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>...
qbus: Make child devices links
Make qbus children show up as link<> properties. There is no stableaddressing for qbus children so we use an unstable naming convention.
This is okay in QOM though because the composition name is expected tobe what's stable....
qbus: Initialize in standard way
Move code to an initfn and finalizer.Replace do_qbus_create_inplace() with qbus_realize().
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>Signed-off-by: Andreas Färber <afaerber@suse.de>
qdev: Remove qdev_prop_exists()
Can be replaced everywhere with object_property_find().
qom: Push error reporting to object_property_find()
Avoids duplicated error_set().
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>[AF: Also drop error_set() in object_property_del().]Signed-off-by: Andreas Färber <afaerber@suse.de>
qdev: Push "type" property up to Object
Now that Object is a type, add an instance_init function and pushthe "type" property from qdev to there.
qdev: Move bus properties to abstract superclasses
In qdev, each bus in practice identified an abstract superclass, butthis was mostly hidden. In QOM, instead, these abstract classes areexplicit so we can move bus properties there.
All bus property walks are removed, and all device property walks...
qdev: Clean up global properties
Now that global properties do not depend on buses anymore, setthem directly in the device instance_init function.
qdev: release parent properties on dc->init failure
While looking into hot-plugging bridges, I can create a qemu segfault via:
$ device_add pci-bridge
Bridge chassis not specified. Each bridge is required to be assigned a unique chassis id > 0. **ERROR:qom/object.c:389:object_delete: assertion failed: (obj->ref == 0)...
qdev: Fix adding of ptr properties
ptr properties have neither a get/set or a print/parse which means that whenthey're added they aren't treated as static or legacy properties.
Just assume properties like this are legacy properties and treat them as such....
qom: Refine container_get() to allow using a custom root
Specify the root to search from as argument. This avoids hardcoding"/machine" in some places and makes it more flexible.
Signed-off-by: Andreas Färber <afaerber@suse.de>Cc: Paolo Bonzini <pbonzini@redhat.com>...
qdev: qdev_unplug(): use error_set()
It currently uses qerror_report(), but next commit will convertthe drive_del command to the QAPI and this requires usingerror_set().
One particularity of qerror_report() is that it knows when it'srunning on monitor context or command-line context and prints the...
qdev: give all devices a canonical path
A strong limitation of QOM right now is that unconverted ports(e.g. all...) do not give a canonical path to devices that arepart of the board. This in turn makes it impossible to replacePROP_PTR with a QOM link for example....
qdev: put all devices under /machine
Avoid cluttering too much the QOM root.
qdev: Fix qdev_try_create() semantics
Since QOM'ification, qdev_try_create() uses object_new() internally,which asserts "type != NULL" when the type is not registered.This was revealed by the combination of kvmclock's kvm_enabled() checkand early QOM type registration....
qom: Unify type registration
Replace device_init() with generalized type_init().
While at it, unify naming convention: type_init([$prefix_]register_types)Also, type_init() is a function, so add preceding blank line wherenecessary and don't put a semicolon after the closing brace....
qdev: let QOM free properties
Drop the special free callback. Instead, register a "regular" release method in the non-legacy property.
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
qdev: initialize properties via QOM
Similarly, use the object properties also to set the defaultvalues of the qdev properties. This requires reorderingregistration and initialization.
qdev: remove direct calls to print/parse
There's no need to call into ->parse and ->print manually. TheQOM legacy properties do that for us.
Furthermore, in some cases legacy and static properties have exactlythe same behavior, and we could drop the legacy properties right away....
qdev: allow reusing get/set for legacy property
In some cases, a legacy property does need a special print methodbut not a special parse method. In this case, we can reuse the get/setfrom the static (non-legacy) property.
If neither parse nor print is needed, though, do not register the...
qdev: split out common init to instance_init
This gets us closer to being able to object_new() a qdev type and have afunctioning object verses having to call qdev_create().
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
qdev: refactor away qdev_create_from_info
Note that the FIXME gets fixed in series 4/4. We need to convert BusState toQOM before we can make parent_bus a link.
qdev: split out UI portions into a new function
qdev-monitor.c deals with the -device, device_add, and info qdm/qtreeinterfaces.
qdev: nuke qdev_init_chardev()
I'm sure the intentions were good here, but there's no reason this should be inqdev. Move it to qemu-char where it belongs.
qom: move properties from qdev to object
This is mostly code movement although not entirely. This makes properties partof the Object base class which means that we can now start using Object in ameaningful way outside of qdev.
qdev: implement cleanup logic in finalize
qdev: remove baked in notion of aliases (v2)
Limit them to the device_add functionality. Device aliases were a hack basedon the fact that virtio was modeled the wrong way. The mechanism for aliasingis very limited in that only one alias can exist for any device....
qdev: kill off DeviceInfo list
Teach the various bits of code that need to walk through available devices todo so via QOM.
Signed-off-by: Anthony Liguori
qdev: kill off DeviceInfo
It is no longer used in the tree since everything is done natively throughQEMU Object Model.
qdev: remove info from class
Now DeviceInfo is no longer used after object construction. All of therelevant members have been moved to DeviceClass.
qdev: allow classes to overload qdev functions
This allows us to drop per-Device registration functions by allowing theclass_init functions to overload qdev methods.
qdev: refactor device creation to allow bus_info to be set only in class
As we use class_init to set class members, DeviceInfo no longer holds thisinformation.
qdev: make DeviceInfo private
Introduce accessors and remove any code that directly accesses DeviceInfomembers.
pci: convert to QEMU Object Model
qdev: use a wrapper to access reset and promote reset to a class method
qdev: add a interface to register subclasses
In order to introduce inheritance while still using the qdev registrationinterfaces, we need to be able to use a parent other than TYPE_DEVICE. Add anew interface that allows this.
qdev: add class_init to DeviceInfo
Since we are still dynamically creating TypeInfo, we need to chain theclass_init function in order to be able to make use of it within subclasses ofTYPE_DEVICE.
This will disappear once we register TypeInfos directly....
qdev: don't access name through info
We already have a QOM interface for this so let's use it.
qdev: move qdev->info to class
Right now, DeviceInfo acts as the class for qdev. In order to switch to aproper ObjectClass derivative, we need to ween all of the callers off ofinteracting directly with the info pointer.
qdev: integrate with QEMU Object Model (v2)
This is a very shallow integration. We register a TYPE_DEVICE but only useQOM as basically a memory allocator. This will make all devices show up asQOM objects but they will all carry the TYPE_DEVICE.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>...
vga: improve VGA logic
Improve VGA selection logic, push check for device availabilty to vl.c.Create the devices at board level unconditionally.
Remove now unused pci_try_create*() functions.
Make PCI VGA devices optional.
Reviewed-by: Jan Kiszka <jan.kiszka@siemens.com>...
qdev: fix device_del by refactoring reference counting
Commit 8eb0283 broken device_del by having too overzealous reference countingchecks. Move the reference count checks to qdev_free(), make sure to removethe parent link on free, and decrement the reference count on property removal....
qdev: fix hotplug when no -device is specified
The peripheral[-anon] containers are initialized lazily but since they sit onsysbus, they can not be created after realize. This was causing an abort() tooccur during hotplug if no -device option was used....
qom: distinguish "legacy" property type name from QOM type name
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
qom: register qdev properties also as non-legacy properties
Push legacy properties into a "legacy-..." namespace, and make themavailable with correct types too.
For now, all properties come in both variants. This need not be thecase for string properties. We will revisit this after -device is...
qom: fix swapped parameters
Reviewed-by: Anthony Liguori <anthony@codemonkey.ws>Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
qom: push permission checks up into qdev_property_add_legacy
qdev_property_get and qdev_property_set can generate permissiondenied errors themselves. Do not duplicate this functionality inqdev_get/set_legacy_property, and clean up excessive indentation....
qom: interpret the return value when setting legacy properties
qom: optimize qdev_get_canonical_path using a parent link
The full tree search was a bit unreasonable.
qom: add string property type
qdev: add a qdev_get_type() function and expose as a 'type' property
qdev: provide a path resolution (v2)
There are two types of supported paths--absolute paths and partial paths.
Absolute paths are derived from the root device and can follow child<> orlink<> properties. Since they can follow link<> properties, they can be...
qom: add child properties (composition) (v3)
Child properties express a relationship of composition.
qom: add link properties (v2)
Links represent an ephemeral relationship between devices. They are meant toreplace the qdev concept of busses by allowing more informal relationshipsbetween devices.
Links are fairly limited in their usefulness without implementing QOM-style...
qdev: add explicitly named devices to the root complex
We first add a 'peripheral' container to the root device that we add usercreated devices to. This provides all user created devices with a unique andisolated namespace.
dev: add an anonymous peripheral container
qom: add a reference count to qdev objects
To ensure that a device isn't removed from the graph until all of its links arebroken.
qom: add new dynamic property infrastructure based on Visitors (v2)
qdev properties are settable only during construction and static to classes.This isn't flexible enough for QOM.
This patch introduces a property interface for qdev that provides dynamic...
qom: register legacy properties as new style properties (v2)
Expose all legacy properties through the new QOM property mechanism. The qdevproperty types are exposed through the 'legacy<>' namespace. They are alwaysvisited as strings since they do their own string parsing....
qom: introduce root device
This is based on Jan's suggestion for how to do unique naming. The root deviceis the root of composition. All devices are reachable via child<> links fromthis device.
qdev: provide an interface to return canonical path from root (v2)
The canonical path is the path in the composition tree from the root to thedevice. This is effectively the name of the device.
This is an incredibly unefficient implementation that will be optimized in...
qdev: Fix crash on -device '?=x'
Spotted by Coverity.
Signed-off-by: Markus Armbruster <armbru@redhat.com>Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
qdev: switch children device list to QTAILQ
SCSI buses will need to read the children list first-to-last. Thisrequires using a QTAILQ, because hell breaks loose if you just tryinserting at the tail (thus reversing the order of all existingvisits from last-to-first to first-to-tail)....